Particle size analysis: A comparison of laboratory-based techniques and their application to geoscience

Abstract

In sedimentary geoscience, the particle size distribution (PSD) of a sediment has a fundamental effect on a sediment's ability to be entrained, eroded, and deposited. Therefore, it is crucial to accurately measure the PSD of sediments. Several laboratory-based methods of particle size analysis are commonly employed in geoscience; however, each method is based on different principles and the comparison of data from one technique to another is challenging. In this study, we have compared the output of four commonly-used laboratory-based techniques: Laser Particle Size Analysis (LPSA), optical point counting, 2D automated image analysis, and X-ray Computed Tomography (XCT). Each technique has been used to measure eight samples of spherical silica particles, all prepared with known particle size ranges. Spherical particles have been used to minimise the effects of variable sorting and particle shape on data output. Here we have compared the differences between the measured PSD and descriptors of each PSD, showing that, at small particle diameters (<150 μm), all techniques agree. However, at particle diameters >150 μm, LPSA overestimates the size of particles, due to limitations in the way that particle diameter is calculated by this technique. In contrast, 2D automated image analysis and optical point counting underestimate the diameters of particles, due to stereology (e.g., the effect of slicing particles during thin section preparation). Results from XCT analyses have the lowest values of sorting (range of measured particle diameters) and are therefore the most tightly constrained. In addition, XCT is the only 3D analysis method, allowing particle shape, orientation, and intraparticle porosity to be measured for a volume of material. We therefore conclude that XCT is the most accurate way to determine a grain size distribution in sediments.